A Reliability Analysis Method of Suspension Bridge based on Support Vector Machines

Abstract eng:
To mostly complex bridges, the performance functions can not be expressed explicitly. The reliability analysis needs times of finite element calculations. It costs much time and it is difficult in calculating the failure probability. Therefore, over the years the reliability of long span suspension bridges was rarely studied and reliability analysis of bridges was mainly for simple structures or components. In this paper, a reliability analysis method based on Support Vector Machine (SVM) was proposed. Uniform design method (UDM) was adopted to select training data and SVM was used as response surface. Structural reliability index was calculated in combination with the traditional reliability analysis methods (such as, the first-order reliability method (FORM), the second-order reliability method (SORM) or Monte Carlo simulation method (MCSM)). Numerical examples show that sampled with the UDM can greatly reduce the number of samples required for training by SVM model, and a very good approximation of the limit state surface can be obtained to get the failure probability. The reliability analysis of the serviceability limit state of a typical self-anchored suspension bridge––Liede Bridge was carried out by the method proposed. The analysis results show that the method is feasible and has a good prospect, which provides a new way for complex structural reliability analysis.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket